• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

[Help Thread] The "Square-1 Help / Alg Sharing" thread

TMarshall

Member
Joined
Feb 13, 2015
Messages
132
Location
Bay Area, CA, USA
WCA
2014MARS04
YouTube
Visit Channel
Feel free to criticize my solves :) this was my second best avg btw :tu

http://youtu.be/r2vm-BoxHNk

A few things: your recognition needs work. When you're executing your eo alg, you should already have a general idea of what your cp case will be.same with ep. I'm not sure if you're doing this in this video, but block preservation helps so much. This will reduce almost all of your eps to combinations of Opp, adj, and u perms. Finally, you should probably just spam tps. When I averaged high 14's, I decided to just spam tps where I could, and it dropped a second off of my times. Hope this helps.
 

Lid

Member
Joined
Jul 8, 2008
Messages
909
Location
Sweden
WCA
2008LIDS01
Been looking into the ODD vs. EVEN flip thing for some EPs lately, here are two more:
W/W: 1,0/3,0/-1,-1/4,1/-1,-1/-2,-2/-1,-1/-2,4/-1,0 [8|23], normal alg is 7w
adj/U: /-3,-3/2,-1/-2,0/-2,0/0,-4/4,0/0,-2/-2,0/-3,0/-3,-3/0,-3 [11|25], normal alg is 10w
 

TMarshall

Member
Joined
Feb 13, 2015
Messages
132
Location
Bay Area, CA, USA
WCA
2014MARS04
YouTube
Visit Channel
Been looking into the ODD vs. EVEN flip thing for some EPs lately, here are two more:
W/W: 1,0/3,0/-1,-1/4,1/-1,-1/-2,-2/-1,-1/-2,4/-1,0 [8|23], normal alg is 7w
adj/U: /-3,-3/2,-1/-2,0/-2,0/0,-4/4,0/0,-2/-2,0/-3,0/-3,-3/0,-3 [11|25], normal alg is 10w

Since I have no clue how to gen algs, and you obviously do, can you gen me some algs for the W-U and U-W cases? If you can, both even and odd cases because I personally don't like the ones that others have found.
 

Lid

Member
Joined
Jul 8, 2008
Messages
909
Location
Sweden
WCA
2008LIDS01
Since I have no clue how to gen algs, and you obviously do, can you gen me some algs for the W-U and U-W cases? If you can, both even and odd cases because I personally don't like the ones that others have found.

Here is a list of some W/U's I compiled:

tomasmu
/-3,-3/0,-3/4,-2/0,-2/0,2/0,4/0,-4/0,2/-1,4/-3,-3/3,6 [11|27] W/U+ (FR/L)
/3,3/1,-4/0,-2/0,4/0,-4/0,-2/0,2/-4,2/0,3/3,3/-3,6 [11|27] W/U- (FL/R)
/3,3/3,0/2,-4/2,0/-2,0/-4,0/4,0/-2,0/-4,1/3,3/6,-3 [11|27] U+/W (L/FR)
/-3,-3/4,-1/2,0/-4,0/4,0/2,0/-2,0/-2,4/-3,0/-3,-3/6,3 [11|27] U-/W (R/FL)

Vandenbergh
/-3,0/4,0/-3,0/0,3/-1,0/-2,0/-4,-4/0,-2/0,1/-4,0/0,3/0,3 [12|25] W/U+ (LB/R)
/0,-3/4,0/0,-1/0,2/4,4/2,0/1,0/0,-3/3,0/-4,0/3,0/0,-3 [12|25] W/U- (FR/F)
/0,3/0,-4/0,3/-3,0/0,1/0,2/4,4/2,0/-1,0/0,4/-3,0/-3,0 [12|25] U+/W (R/LB
/3,0/0,-4/1,0/-2,0/-4,-4/0,-2/0,-1/3,0/0,-3/0,4/0,-3/3,0 [12|25] U-/W (F/FR)

"Chinese list"
/3,3/-1,0/-2,0/-4,0/-4,0/0,-4/0,-2/-2,2/-1,-1/-3,0/3,3/-3,6 [12|29] W/U+ (RB/F)
/3,3/5,0/-2,0/-4,0/-4,0/0,-4/0,-2/-2,2/-1,-1/3,0/3,3/-3,0 [12|28] W/U- (LB/F)
/3,3/ 1,0/-2,0/-4,0/0,-4/0,-4/0,-2/2,-2/1,1/3,0/3,3/6,3 [12|29] U+/W (F/RB)
/3,3/-5,0/-2,0/-4,0/0,-4/0,-4/0,-2/2,-2/1,1/-3,0/3,3/0,3 [12|28] U-/W (F/LB)

"Crawford list"
/3,3/-3,0/-2,4/-2,0/2,0/4,0/-4,0/2,0/1,-4/-3,-3/0,3 [11|26] W/U+ (LB/R)
/3,3/-3,0/-2,4/0,-2/2,0/0,4/-4,0/0,2/1,2/-3,-3/-3,6 [11|27] W/U- (LB/B)
/3,3/-3,0/-4,2/2,0/-2,0/-4,0/4,0/-2,0/-4,1/3,3/3,6 [11|27] U+/W (F/FL)
/3,3/-3,0/-4,2/0,2/-2,0/0,-4/4,0/0,-2/-1,-2/-3,-3/3,0 [11|26] U-/W (R/FL)

= /alg/ [w|m] (where op edges are/where solved bar in U is)

The only alg I have generated so far: (did this some years ago)
/-3,-3/0,-1/-2,0/-2,0/-2,-2/0,2/2,0/-2,0/2,0/-1,-2/-3,-3/-3,0 [12|28] U+/W (R/LB)

Optimal length seem to be 11 twists or 24 moves.
 
Joined
Apr 23, 2010
Messages
1,391
Location
Scotland, UK
WCA
2009SHEE01
YouTube
Visit Channel
I'm not overly interested in doing parity and OBL in inspection, which some people have discussed as a theoretical possibility, but after some thought I suppose it might be fun to look into whether it could work out. I don't expect to work on it much, but if someone could give me some indication of what would actually be useful it would help. I don't use OBL (or even CO->EO->..., though I do know how to do it), so some in depth discussion/video from someone who looks for OBL in inspection is what I would really need.
 

TMarshall

Member
Joined
Feb 13, 2015
Messages
132
Location
Bay Area, CA, USA
WCA
2014MARS04
YouTube
Visit Channel
So I was average low 12's on sq1, with a pb average of 100 of 12.01. I had been stuck at this for a few weeks or so, and then yesterday, I suddenly got an 11.44 average of 100. So basically, does this happen to anyone else? I've had a few big jumps like this, and this really only happens to me on sq1.
 

Sam N

Member
Joined
Jun 9, 2015
Messages
89
I'm not overly interested in doing parity and OBL in inspection, which some people have discussed as a theoretical possibility, but after some thought I suppose it might be fun to look into whether it could work out. I don't expect to work on it much, but if someone could give me some indication of what would actually be useful it would help. I don't use OBL (or even CO->EO->..., though I do know how to do it), so some in depth discussion/video from someone who looks for OBL in inspection is what I would really need.

I'm guessing that notkevin is the best person to ask for this. He seems like one of the people who knows the method quite well, however I can't speak directly for him.
 

not_kevin

Member
Joined
Jul 23, 2008
Messages
495
Location
California, USA
WCA
2008YOUN02
YouTube
Visit Channel
I'm not overly interested in doing parity and OBL in inspection, which some people have discussed as a theoretical possibility, but after some thought I suppose it might be fun to look into whether it could work out. I don't expect to work on it much, but if someone could give me some indication of what would actually be useful it would help. I don't use OBL (or even CO->EO->..., though I do know how to do it), so some in depth discussion/video from someone who looks for OBL in inspection is what I would really need.

Unfortunately, I can't help all that much here, because I don't quite combine OBL into everything past simple tracing. However, for a decent number of cases, I'll predict corners because I've memorized which layer they end up in - but because I haven't memorized where exactly the corners go, this is mostly only useful for CO skips and preservation when I know I can do cubeshape in particular ways (the most common being that I don't do anything more than a 1 on bottom during cubeshape if I can help it).

One very very big difference between how you do parity+cubeshape and how I currently do my partial cubeshape + OBL is that I don't have a "fixed" way to do many cases, because I'm always looking to preserve something, get some lucky skips, or simply make cubeshape more ergonomic. Even something as easy as / 0,3 I can optimize, by doing z2 -3,0 /. This is a big reason why I haven't memorized where pieces will end up consistently, because I don't have many cases that have a fixed solution - the only one I can think of having a really consistent solution is kite/square and really short cases (in which I try to just trace out the OBL that results).

I'm not sure what the best way to merge both parity and OBL into inspection would be, because although both rely on the same metadata (the permutation of the pieces in cubeshape), they both only use a piece of it, and they use different pieces (parity only requires a count for the number of 2-swaps, while OBL doesn't need to know yellow/white permutation information). I do think that parity+CS -> OBL -> PBL is the fastest of the 3-step method ideas that exist at the moment, because eliminating parity with an average of like 1 slice is pretty great, but my efforts to properly learn PCS have been stymied by my old age and dedication to other commitments =/

I'll try to think about this more, 'tho, and hopefully come up with some ideas for doing CS+P+OBL, now that I know how you handle CS+P in inspection.

EDIT: I figured it'd be a good idea to have some examples in how I influence OBL during CS, so people have a better idea of what I mean - after posting, I realized that most people probably don't really do this, so it's difficult to explain to them :p

Scramble: (-2, -3) / (-3, 0) / (3, 0) / (0, -3) / (5, -4) / (6, 0) / (0, -5) / (-3, 0) / (0, -3) / (-4, 0) / (4, -4) / (3, 0) / (-2, 0) / (-1, 0)
CS: 0,2 / 3,-4 / -1,-2 / 0,-3 /
Explanation: I go to kite-scallop with this case, but because kite-scallop is one of those nice cases that can be done pseudo-2gen (in which one layer only moves at most 1 per slice), I know that if I do 0,2 / 3,-4, I'll be set up. The fingertricking is significantly slower than the "conventional" approach of doing 0,2 / -3,2 / 1,2 / 0,3 /, but I get quite a bit of OBL knowledge.

Scramble: (-2, 3) / (5, -1) / (0, -3) / (3, 0) / (1, -5) / (3, 0) / (0, -4) / (3, 0) / (0, -3) / (-2, -4) / (4, 0) / (2, -2) / (0, -3)
CS: y2 2,-3 / 4,0 / 2,0 / 1,0 / -4,0 / 3,0 /
Explanation: similar to the above, I see that doing -4,3 / 4,0 / sets up into the 2-gen paw-paw case (or whatever the name of that case is), and doing so preserves the nice block in DL. I did y2 2,-3 / because I felt it was nicer to fingertrick.

Scramble: (0, 5) / (-3, 6) / (4, -5) / (6, -3) / (2, -3) / (-3, 0) / (4, -1) / (2, -2) / (-2, 0) / (-4, 0) / (-2, 0) /
CS: y2 2,0 / -4,1 / -1,0 / -3,0 /
Explanation: a combination of tricks above with more prediction based on experience. I see that 0,-2 / goes into shield/square, which I know can be done 2-gen; because of that, and because the UR half already has a corner-edge-corner block ready, I figure doing a y2 will be simpler. Then, after more inspection, I see that although I can't get a CO skip (which happens when the square has an adjacent CO pair, and the shield also has an adjacent CO pair), by doing y2 2,0 / -4,1, the yellow corner-edge block will actually stay together and go to the D layer. Further, because I'm doing y2 2,0 / [whatever],1, see that I'm also going to get a full yellow half on DL, which means that now I know that the D layer is guaranteed to be a 3/4 layer, with the "incorrect" white corner-edge pair in DBR and DR. I don't know what's on top, unfortunately, but this means that all I need to do is check what's on top to do OBL from here. (Conveniently enough for me, that OBL turns out to just be -5,0 / 3,0 / :p)
 
Last edited:
Joined
Apr 23, 2010
Messages
1,391
Location
Scotland, UK
WCA
2009SHEE01
YouTube
Visit Channel
@not_kevin: Hmm ... thanks. I thought you were doing a little more, but I really just wanted to know what your process was so it's still helpful. I'll have a think about it and if/when I get something worthwhile I'll post about it. Is there any info on "nice" OBL cases? I downloaded the pdf of all the cases by depth, so maybe <=x twists is good for x=3 or x=4?

Also, eww cube rotations on sq-1. Is that normal or am I just a noob for not doing them?
 

not_kevin

Member
Joined
Jul 23, 2008
Messages
495
Location
California, USA
WCA
2008YOUN02
YouTube
Visit Channel
@not_kevin: Hmm ... thanks. I thought you were doing a little more, but I really just wanted to know what your process was so it's still helpful. I'll have a think about it and if/when I get something worthwhile I'll post about it. Is there any info on "nice" OBL cases? I downloaded the pdf of all the cases by depth, so maybe <=x twists is good for x=3 or x=4?

Also, eww cube rotations on sq-1. Is that normal or am I just a noob for not doing them?

I'm hoping that I can get some time in to start making some kind of OBL tutorial, but I would say anything thats <=2 twists is lucky, 3 twists is good, 4 twists is nice if the case itself is nice (eg, pseudo-2gen, goes into M2, etc.), and more than that is bad. Since 6 is the max, and there are only really 2 cases (both of which are 1-slice CO cases), the bad cases are roughly 4-6, although I'm a fan of a decent number of the 4-move cases.

The only cases I can consistently get all the way to OBL are the super easy cubeshapes (<= 2 slices, and some 3-slice cases), and that's just because I can trace it to completion rather than because I've memorized anything. However, the way I'd likely do CS+OBL would be to memorize how pieces move for all 3-slice cases and maybe some 4s, then basically trace the extra 1 or 2 moves of the shape I'm in to a case I know and go from there.

Regarding rotations, I'm actually not sure, because most of the other top-tier sq1-ers (or whatever the term is - we really need a better term for this, guys :p) don't rotate nearly as much as I do. But I figure that any advantage I get is a good one, and since rotations don't do anything other than change what color's in front, I don't think there's a significant slowdown for the final ABF.
 

Sam N

Member
Joined
Jun 9, 2015
Messages
89
@not_kevin: Hmm ... thanks. I thought you were doing a little more, but I really just wanted to know what your process was so it's still helpful. I'll have a think about it and if/when I get something worthwhile I'll post about it. Is there any info on "nice" OBL cases? I downloaded the pdf of all the cases by depth, so maybe <=x twists is good for x=3 or x=4?

Also, eww cube rotations on sq-1. Is that normal or am I just a noob for not doing them?

If you're referring to rotations done before the solve during inspection, I would say they're pretty common, at least for me anyway. I see no reason to not do them, they let you get a better angle before you solve. Rotations during solves however, I've not seen anyone do that. But I'm pretty sure that's been established as being very ineffective.
 
Top